Prevent and Conquer
A collection of key research updates that headlined this year's Conference on Retroviruses and Opportunistic Infections
By Kristen Jill Kresge and Richard Jefferys*
The success of antiretroviral (ARV) treatment is a remarkable victory in the now 29-year-old battle against HIV. In his plenary talk at the 17th Conference on 
Retroviruses and Opportunistic Infections (CROI), which was held from February 16-19 in San Francisco, Anthony Fauci, director of the National Institute of Allergy and Infectious Diseases (NIAID), called ARV therapy “one of the best success stories in biomedical research as it gets translated to the patient.”
At the end of 2008, approximately four million HIV-infected people in low- and middle-income countries were receiving ARVs. “As the years go by we’re doing better in getting drugs to people who need them,” said Fauci. “That’s the good news.”
“The sobering news,” Fauci continued, “is that it’s not sustainable.” Despite this progress, there is still a huge gap between the number of people who need treatment and those who receive it—a gap recently widened by the World Health Organization’s (WHO) decision to revise treatment guidelines in response to mounting evidence for the benefits of earlier initiation of ARV therapy (see Everything from Cause to Cure, IAVI Report, July-Aug. 2009). Based on the updated guidelines, only 30% of the HIV-infected people in the world who qualify for therapy are receiving it.
Closing this gap is a huge priority in battling HIV/AIDS, and critical to that is reducing the number of new infections. Fauci outlined a triumvirate of HIV prevention strategies that top the research agenda at NIAID, including the development of a preventive HIV vaccine; test and treat, which calls for universal HIV testing and immediate treatment for those infected; and pre-exposure prophylaxis, which involves ARVs delivered orally or in a microbicide gel to uninfected individuals. Several research updates on these three areas dominated the discussions at this year’s CROI. Another priority at NIAID is finding a cure for HIV infection. Fauci said all of these efforts relate to a common goal, which is “controlling and ultimately ending the HIV/AIDS pandemic.”
Building on RV144
One of the main planks in the HIV vaccine research agenda at NIAID is building on the results of the Thai trial, which showed that a prime-boost regimen (a canarypox vector prime followed by an HIV protein boost) provided about 31% protection against HIV infection (1). Building on these results will involve further clinical research, as well as more basic discovery research to support a better understanding of vaccine design and development.
With regard to clinical development, Fauci said the RV144 results may have shifted the focus. “We really have to focus future trials on the prevention of acquisition,” said Fauci. “Understanding the T-cell response [through trials of candidates that are solely designed to impact viral load] is very important, but when we do a large clinical trial in humans, it is my opinion that we’ve really got to look at acquisition.” Fauci also set a high bar for the target efficacy for an HIV vaccine. “I think we’ve got to do better than 60%-70%,” he said, suggesting dramatic improvements on the Thai trial results would be necessary. “We’re setting the bar very high, but the history of AIDS tells us we’ll clear that bar with the best minds, resources, and political will.”
One obvious way to improve upon the RV144 results is to try to determine the immune correlates of protection. Nelson Michael, director of the US Military HIV Research Program (MHRP), who provided an update on some of the post-hoc analyses of RV144 (see Update on RV144, below), said work on trying to determine the correlates of protection is just getting underway, with results expected in a year. Due to the limited number of samples available and the overall low HIV incidence rates during the trial, determining the correlates will likely be difficult. “We may not know what the correlates of protection are,” said Fauci, “but we will know, and this is really important, what they’re not.”
Many researchers have all but ruled out broadly neutralizing antibodies as the correlate of protection in RV144, since neither candidate induced such responses in previous trials. “[The Thai trial] tells us you can prevent acquisition without traditional neutralizing antibodies,” said Fauci.
If correlates of protection can’t be identified from RV144, it is possible that additional efficacy trials will be conducted. And as Michael pointed out, this will require other funding organizations, such as NIAID and the Bill & Melinda Gates Foundation, to pitch in. “The Army could bring US$8 million a year to any further study we would probably do,” said Michael, who noted that the most expensive year of the Thai trial cost $17 million. “We can’t go rogue and do another study by ourselves.”
| Update on RV144 |
|
Nelson Michael, director of the US Military HIV Research Program, opened the session on HIV vaccines at the 17th Conference on Retroviruses and Opportunistic Infections by summarizing the results of RV144, the first AIDS vaccine trial to show any efficacy (1; see Raft of Results Energizes Researchers, IAVI Report, Sep.-Oct. 2009). Subsequently, investigators have conducted a post-hoc evaluation of risk behaviors of the trial participants, as reported at six-month intervals. Before sharing the results, Michael issued this disclaimer: “Your eyebrows should raise anytime any trialist talks about a post-hoc analysis.” He then showed that having reported high-risk behavior at any time during the trial, significantly impacted vaccine-induced protection (p value of 0.008). Michael suggested the association of risk with lack of protection may have more to do with the seemingly transient impact of vaccination—the protective effect of vaccination appears to have been concentrated almost entirely during the first 12 months of the three-year trial—than with risk, since the reporting of risk behaviors continued throughout the trial. Michael also presented data on the titers of binding antibody against gp120 among the vaccinated volunteers. Comparing the magnitude of these responses at two and 24 weeks after the final immunization, Michael demonstrated that they dropped precipitously; the average geometric mean titer of binding antibody against the recombinant CRF01_AE gp120 was 14,558 at the two-week time point and 1,000 at 24 weeks.—RJ |
Studies of neutralizing antibodies
Other HIV vaccine research efforts are focused on studying broadly neutralizing antibodies against HIV, with the ultimate goal of reverse engineering vaccine immunogens capable of inducing these antibodies. However, a provocative presentation by Julie Overbaugh, a member of the Human Biology Division at the Fred Hutchinson Cancer Research Center, raised questions about whether neutralizing antibodies can actually block acquisition of HIV. “We really don’t know yet if antibodies provide some benefit in protecting against HIV in exposed populations,” said Overbaugh.
In nonhuman primates, studies have shown that passive administration of the broadly neutralizing antibody b12 can block infection with a simian immunodeficiency virus/HIV (SHIV) challenge. The b12-treated macaques eventually become infected, but only after substantially more challenges than control animals (2). However, the challenge virus used in these studies was the most neutralization sensitive to b12, according to Overbaugh. “These experimental models provide important proof of principle that these approaches can work, but they focus on optimal combinations of viral strain and neutralizing antibodies. It’s hard to extrapolate that such levels of antibodies with that specificity would neutralize most circulating strains of HIV. The real world is not so ideal,” she said.
To see whether neutralizing antibodies present at the time of exposure could protect against infection in the natural setting, Overbaugh and colleagues studied HIV-infected mothers who breast feed their infants. Overbaugh presented data from an unpublished study of 100 infants, who were HIV uninfected at birth, whose mothers had broad and potent neutralizing antibodies. The maternal antibodies were passively transferred to the infants and were present at birth and during the breast-feeding period. Serum samples from the infants showed that some had “very broad and potent neutralizing antibodies that were passed on from their mothers,” according to Overbaugh. But when she compared the neutralizing antibodies in the 32 infants who acquired HIV during the period of breast feeding to those who remained uninfected, she found there was no difference in their breadth or potency. “In the setting of mother-to-infant transmission there’s no evidence that a broad and potent neutralizing antibody response as measured against a heterologous panel of viruses provided protection from HIV infection,” she said.
Overbaugh said the results might be different with the broadly neutralizing antibodies like PG9 and PG16, which were described in 2009, since they are able to neutralize many more viruses. Fauci said this is why passive immunization studies with some of the newer broadly neutralizing antibodies should be conducted.
Evolution of neutralizing antibodies
Previous research on the evolution of neutralizing antibodies in HIV-infected people has shown that although these antibodies emerge several months after HIV infection, they are unable to neutralize contemporaneous HIV isolates from the same individual (3). Frederick Hecht from the University of California in San Francisco, who presented data from a study addressing the long-term evolution of antibody responses to the shape-shifting HIV envelope in people with HIV infection, found that instead, the antibodies seem to be playing a constant game of catch-up—they are only capable of neutralizing autologous viruses sampled three to four months previously.
Hecht showed that this phenomenon continued for close to four years in the majority of the 13 study participants, with antibody responses continuing to evolve, but always lagging behind the virus.
But when Hecht analyzed the neutralizing activity of these antibodies from the last sample collected in the four-year study, he found that all 13 participants possessed antibodies capable of significantly inhibiting a panel of 12 heterologous HIV isolates. Depending on the concentration cut-off, an average of 50-87% of the panel could be neutralized. Hecht and colleagues hypothesized that the generation of new waves of neutralizing antibodies against autologous virus would correlate with the broadening of activity against the heterologous virus panel. In a multivariate model including CD4 count, viral load, duration of infection, and the cumulative evolution of autologous neutralizing antibodies, this indeed proved to be the case—only the latter factor significantly predicted the breadth of heterologous neutralization, with a p value of <0.0001.
The rationale for this study was partly derived from a study by Hecht’s colleague Larry Bragg, who presented a poster at CROI (see Superinfection Risk May Decline Over Time, below) indicating that one year after acquisition of HIV infection, the risk of superinfection drops significantly. Hecht concluded by suggesting that the time course of antibody evolution documented in the study was consistent with the data on decreased risk of superinfection, and that this may represent a model of protection with important implications for vaccine design.
| Superinfection Risk May Decline Over Time |
|
Larry Bragg, a research associate at the Gladstone Institute of Virology and Immunology at the University of California in San Francisco, presented a poster suggesting that among people recently infected with HIV, the risk of superinfection with a second virus strain declines significantly after one year. The data were obtained from the OPTIONS study and involved 256 individuals with recent HIV infection, nine of whom were found to have acquired a highly divergent second strain of HIV. Of these cases, six occurred in the first year, one in the second, and two in the third. Although the numbers are small, statistical analyses indicated an approximately 21-fold reduction in risk of superinfection after the first year (p value of 0.005). Bragg and colleagues suggested that biological mechanisms may account for the finding (see Evolution of neutralizing antibodies, above), but also acknowledged that other explanations are possible, such as variations in individual susceptibility. —RJ |
Locking onto lipids
In 2001 it was reported that persistent infection with GB virus type C (GBV-C) was associated with prolonged survival in people with HIV (4). GBV-C is a flavivirus that is not known to cause disease, and the mechanism by which it might impact HIV infection has remained shrouded in mystery. At CROI, Heide Reil, from the University Hospital in Erlangen, Germany, presented data showing that, surprisingly, some antibodies directed against the E2 glycoprotein of GBV-C can neutralize a wide range of different HIV isolates even though there is no homology between the two viruses. Reil’s investigation was prompted by data in the 2001 NEJM paper suggesting that people with HIV who have cleared GBV-C and possess anti-E2 antibodies still show attenuated disease progression, even though the magnitude of the effect was reduced compared to individuals who have persistently detectable GBV-C viremia.
Reil described initial experiments conducted with a polyclonal mix of different monoclonal antibodies directed against E2, which revealed broad neutralization of a panel of HIV isolates from multiple clades, including C and D, which are typically hard to neutralize. Further analyses showed that just two of the monoclonal anti-E2 antibodies, dubbed M6 and M11, could recapitulate this activity on their own. However, neither could neutralize other viruses tested, including murine moloney virus, yellow fever, and adenovirus serotypes 5 and 12.
Intrigued by these observations, Reil and colleagues embarked on an iterative series of studies that revealed the neutralizing activity was not directed against HIV envelope glycoproteins, but rather phospholipids contained in the viral membrane. This class of lipids inhabits the inner layer of cell walls and they are acquired by HIV particles as they bud. Reil noted that two broadly neutralizing monoclonal antibodies against HIV, 2F5 and 4E10, are also known to bind phospholipids. In those cases, a mechanism has recently been suggested which involves a two-step process: the antibodies initially bind phospholipids on the viral membrane, which then allows them to be in close proximity when the epitope they target in HIV envelope’s membrane proximal region (MPER) is very briefly exposed during fusion of the virus to the target CD4 cell (see Figure 4, below). This recently published mechanism was described at CROI in a separate talk by Bing Chen (5).
| Figure 4 |
|
|
To assess whether the M6 and M11 antibodies were targeting the same phospholipids, Reil tested their binding to an array of different types, finding that they bound a phosphatidylinositolphosphate called PI(4)P2 (see Targeting Phospholipids, below). In contrast, 2F5 and 4E10 bound to cardiolipin, and cardiolipin and phosphatidylserine respectively, as has previously been reported (6). Reil concluded her talk by suggesting that the E2 glycoprotein, and perhaps even entire GBV-C particles, should be considered as antigens for future HIV vaccine strategies.
| Targeting Phospholipids |
|
In a plenary talk at CROI, Hans-Georg Kräusslich, a professor at the University of Heidelberg, reported data suggesting that HIV specifically recruits phosphatidylinositolphosphate PI(4,5)P2, when assembling the viral membrane (see Locking onto lipids, above). His group is now working to understand the process in more detail. Commenting on whether it might be possible to target HIV using this knowledge, Kräusslich said, “I don’t know but there is at least some promising reason to study it.” Targeting phospholipids may not carry the risk of autoimmunity typically associated with host cell components, because one of their functions is to promote non-inflammatory clearance of apoptotic cells (7). —RJ |
Live-attenuated protection
For nearly two decades, researchers have been trying to unravel the mechanisms underlying the protection afforded by live-attenuated SIV vaccines in rhesus macaques. At CROI, Paul Johnson, chair of the division of immunology at the New England Primate Research Center, reviewed data from multiple challenge experiments using the attenuated vaccine SIVmac239Δnef, developed by Johnson’s colleague Ron Desrosiers. These studies have consistently shown sterile protection against homologous SIV isolates administered intravenously. And although protection against a high-dose mucosal challenge with the heterologous SIVsmE660 has not been achieved, immunized animals display enhanced post-infection control of viremia. Johnson also cited a recent unpublished study conducted by Matt Reynolds involving repeated low-dose mucosal challenges that showed this live-attenuated vaccine resulted in a significant delay in time of acquisition of SIVsmE660, along with control of viral load in animals that became infected.
Johnson described several studies conducted to delineate the contribution of antibodies to the protection observed with SIVΔnef. The first, by the research group of Jörn Schmitz at Beth Israel Deaconess Medical Center in Boston, administered the B cell-depleting monoclonal antibody rituximab to 10 vaccinated macaques for 100 days prior to challenge with SIVmac239, leading to persistent depletion of B cells in five of the animals. Despite the absence of B cells and antibody responses, four of five were completely protected and the remaining animal has controlled viremia through 200 days post-challenge.
In another unpublished study summarized by Johnson, Ron Desrosiers employed a different approach that cleverly took advantage of the poorer protection SIVΔnef offers against SIVsmE660. Desrosiers swapped the SIVmac239 envelope gene from SIVΔnef with another derived from SIVE543, which is closely related to SIVsmE660. Immunogenicity analyses have uncovered no cross-neutralizing antibodies against these two envelopes. An experiment was then conducted in which 24 macaques were divided into two groups and received either SIVΔnef or SIVΔnef/EnvE543. An SIVmac239 challenge virus was also created with the envelope of SIVE543 and six animals in each group were challenged with wild type SIVmac239 and six with SIVmac239/EnvE543. This approach allowed each group to receive both homologous and heterologous challenges. No vaccinees in either group challenged with homologous virus were infected, and only two of 12 that received a heterologous challenge were infected. Considering the data from Desrosiers and Schmitz, Johnson concluded that antibody responses are unlikely to play a major role in the protection observed with live-attenuated vaccines.
Johnson next explained that ongoing antigenic stimulation by SIVΔnef does appear to play a role in protection, citing data using a single-cycle SIV that, as its name suggests, can only complete one round of infection. Published studies using this construct as a vaccine have shown only around a log reduction in post-challenge viral load compared to the complete protection offered by live-attenuated SIV (8).
In the last section of his talk, Johnson focused on the role of CD8+ T-cell responses in explaining the efficacy of SIVΔnef. To address this issue, his laboratory performed a study in which three groups of macaques were challenged intra-vaginally with SIVmac251 at different timepoints after immunization. Each group consisted of nine animals, six that received SIVΔnef and three controls. The challenge was administered at 5, 20, or 40 weeks after vaccination. No protection was seen at five weeks, whereas three of six macaques challenged after 20 weeks were protected, as were two at week 40. All but one of the nine controls became infected.
A starker contrast emerged when he looked at post-challenge viremia in these animals. Although the magnitude of SIV-specific CD8+ T-cell responses peaked at five weeks, the group challenged at this timepoint showed a log reduction in peak viral load compared to the controls. However, the difference between the two groups was extremely short-lived. In contrast, both the 20- and 40-week groups showed far greater reductions in viral load that remained more than two logs lower than controls through about 40 weeks.
A comprehensive analysis of SIV-specific CD8+ T-cell responses revealed that expression of the central memory markers CCR7 and CD127 (also referred to as IL-7R, the receptor for IL-7) was significantly higher at 20 and 40 weeks. But Johnson cautioned against leaping to the conclusion that this CD8+ T-cell phenotype was the cause of enhanced control of viral load, as similar expression levels of these molecules have been reported after single cycle SIV immunization. The marker that distinguished the two approaches was programmed death (PD)-1, once thought to signal T-cell exhaustion, but now known to also be upregulated on activated T cells, leading Johnson to suggest that in SIVΔnef vaccinated animals its expression was serving as an indicator of persistent antigenic stimulation.
Johnson proposed what he called the Goldilocks principle: “You need just the right amount of ongoing antigenic stimulation,” he said, “too much and all of your cells end up as effector memory cells, too little, they turn into central memory cells.” His opinion, at least in these models, is that SIVΔnef seems to be just right.
| Researchers Still Grapppling With Transmission |
|
As early as 1992, researchers observed that a single HIV founder virus establishes infection in cases of heterosexual transmission. George Shaw, a professor at the University of Alabama at Birmingham, who delviered the Bernard Fields Memorial Lecture at this year's 17th Conference on Retroviruses and Opportunistic Infections (CROI), has confirmed in much larger cohorts from acute HIV infection studies that a single founder virus establishes infection in 80% of heterosexual transmissions (see HIV Transmission: The Genetic Bottleneck, IAVI Report, Nov.-Dec. 2008). But at CROI, Barbara Felber, chief of the human retrovirus pathogenesis section at the National Cancer Institute, presented data from a study in rhesus macaques that suggests researchers may be missing some of the transmitted viruses that only begin replicating at detectable levels in the chronic phase of infection. In this study, six rhesus macaques were infected either intra-rectally or intra-vaginally with SIVmac251, a swarm of virus variants with a maximum divergence of 2.3% in the entire Env region and 4% in the V1/V2 loops. The viral loads in the monkeys peaked at 7.5 log HIV RNA copies/ml of blood during the acute phase of infection (two weeks after infection) and 5.8 log RNA copies/ml in the chronic phase of infection (30 weeks post-infection). Felber then sequenced the virus variants present at both stages of infection using single genome amplification. The power calculations employed were capable of identifying any genetic variants that represented at least 5% of the total virus population. In three of the six animals, Felber found that the genetic sequences present during acute infection fit into a “nice Poisson distribution,” suggesting they were infected by a single founder virus. The founder virus, however, was not the same, indicating that multiple variants in the challenge stock could establish infection. In chronic infection, Felber observed that two of these three animals had some virus variants that could not have evolved from the acute virus. In two other animals, two virus variants were detected in the acute phase. Both animals also had additional variants present in the chronic phase. In total, five of six animals had multiple variants in the chronic phase that were not detected during acute infection. Felber said that multiple founder viruses cross the mucosal barrier, and while only one or two of these replicate at detectable levels during the acute phase, other variants may emerge to prominence later, during the chronic phase. She suggested that only estimating the number of transmitted virus variants in humans during the acute phase may be inaccurate. —KJK |
The test and treat approach
When researchers at the WHO published results from a mathematical model showing that universal annual testing and immediate ARV treatment for all HIV-infected individuals could make a major dent in the number of new HIV infections, the model ignited discussions among researchers and stimulated new research into the feasibility of the so-called test and treat approach (9). A NIAID-funded pilot study of test and treat in Washington, D.C. and New York City is scheduled to start in a few months (see Vaccine Briefs, IAVI Report, Sep.-Oct. 2009).
Despite the attention test and treat has received, there is limited data to support the premise. “All of the mathematical models assume much lower HIV transmission rates when on ARV therapy but there is very little empiric evidence,” said Deborah Donnell, deputy director of the HIV Prevention Trials Network (HPTN) statistical center.
This is beginning to change. At CROI, Donnell presented data that helps bolster the connection between ARV treatment and prevention. In an observational sub-study of the partners in prevention study—a study that showed treatment of herpes simplex virus-2 did not reduce risk of HIV infection—Donnell and colleagues analyzed HIV transmission rates among 3,381 serodiscordant couples from seven countries in eastern and southern Africa.
At the start of the study, the HIV-infected partners, the majority of whom were female (68%), had CD4+ T-cell counts greater than 250 and were not already taking ARVs. CD4 levels in the infected partners were measured every six months. During the study, 10% (349) of the HIV-infected partners initiated ARV treatment. Half of those that started therapy had CD4+ T-cell counts below 200 cells/microliter of blood.
At the end of the two-year study, 151 new HIV infections had occurred. Researchers confirmed, by viral sequencing, that 108 of the infections were linked to the HIV-infected partner in the study. Of the 103 infections included in the final analysis, only a single one occurred when the HIV-infected partner was on ARV treatment. This correlates to a statistically significant 92% reduction in HIV transmission if the infected partner was on ARVs. “There was a substantial prevention benefit for ARV therapy,” said Donnell.
Investigators observed that the infected partners who were not taking ARVs were much more likely to transmit HIV at lower CD4+ T-cell counts. However, even at higher CD4+ T-cell levels, there were still a higher number of HIV infections that occurred between couples if the infected partner was not on ARVs. “HIV transmission occurred across all CD4+ T-cell levels,” Donnell said. Still, she stressed that the highest priority is getting individuals with the lowest T-cell counts on therapy.
An ongoing study, HPTN 052, is a randomized, five-year, Phase III clinical trial designed to determine HIV transmission rates among 1,750 serodiscordant couples in which the infected partner begins ARV therapy immediately, or only when their CD4+ T-cell count falls between 200 and 250 cells/microliter. Researchers hope this study will provide a conclusive answer about the protective effects of ARV therapy.
Another method researchers are employing to gauge the ability of ARV treatment to reduce HIV transmission rates is estimating the community viral load—the mean viral load of all HIV-infected individuals in a given community. And in some cases, declines in community viral load are correlated with declines in the number of individuals newly diagnosed with HIV.
Moupali Das-Douglas, director of the research unit at the San Francisco Department of Public Health, presented data indicating that a significant 40% decrease in the community viral load among men who have sex with men in San Francisco that occurred between 2004 and 2008, correlated with a 45% reduction in the number of new HIV infections during this same four-year period. Additionally, from 2006-2008, Das-Douglas reported that the estimated HIV incidence, which reflects the number of new HIV infections both diagnosed and undiagnosed, also decreased by 33%; however, this was not a statistically significant reduction.
The declines in community viral load and the number of new HIV diagnoses were credited to an increase in HIV testing rates in San Francisco, as well as an increase in the number of infected individuals who are receiving ARVs. By 2008, 90% of HIV-infected individuals with AIDS in San Francisco were receiving ARV treatment. But although there is a statistically significant correlation, this study did not show directly that testing or more widespread ARV treatment was actually what caused the decline in the number of new HIV diagnoses. Still, Das-Douglas concluded that public health departments should consider measuring community viral loads. “What gets measured gets managed,” she said.
Julio Montaner, director of the British Columbia Centre for HIV/AIDS and a proponent of test and treat, reported similar results from a prospective study in British Columbia, Canada, which evaluated the community viral load of all HIV-infected people in the province who are on ARV treatment. Montaner said the rapid uptake of HAART in this population is “driving down viral load steadily,” and that this has resulted in a decrease in the number of new HIV diagnoses, particularly among injection-drug users (IDUs).
In 2004, approximately 50% of IDUs in British Columbia had viral loads greater than 1,500 copies of HIV RNA/ml of blood, compared to fewer than 20% in 2009. Even though the number of HIV tests administered in the study population has risen steadily, according to Montaner, there was a 50% decrease in the number of new HIV diagnoses among IDUs over this five-year period. He credited this reduction in new diagnoses among IDUs to the expansion of ARV therapy in this community, rather than any behavioral modifications. He says there is “plenty of room to improve” on these results, and he hopes to move ahead with what he calls the “seek, test, treat, and retain” strategy with new support from the Canadian government. “We should proceed expeditiously.”
A third study conducted in Washington, D.C., which has the highest HIV prevalence in the US with about 3% of the population living with HIV/AIDS, showed a different trend than what was observed in San Francisco and British Columbia. Researchers from George Washington University School of Public Health and Health Services reported an increase in the number of new HIV diagnoses following a dramatic expansion of routine HIV testing services and efforts to provide those infected with treatment.
A routine HIV testing campaign began in Washington, D.C. in 2006. That year, approximately 35,000 HIV tests were conducted. By 2009, 93,000 HIV tests were performed. And the data suggests that efforts to link people with treatment were successful. In 2004, 23% of HIV-infected individuals in D.C. were entered into an HIV treatment and care program more than 12 months after receiving their HIV diagnosis. This number dropped to 5.4% in 2008. The median CD4+ T-cell count of HIV-infected individuals when they started taking ARVs also increased from 216 cells/microliter of blood in 2004 to 343 in 2008. Additionally, the percentage of people who progressed to AIDS within 12 months of their HIV diagnosis dropped from nearly 50% in 2004 to 28% in 2008.
Yet Amanda Castel, an assistant research professor at George Washington University, reported that from 2004 to 2007 there was a 17% increase in the number of HIV diagnoses in Washington, D.C., from approximately 1,100 to 1,300.
Kimberly Smith, an associate professor at Rush University Medical Center in Chicago, said that late diagnoses are partly responsible for the increased mortality from HIV within the black community in the US. “Black individuals are much more likely to die from HIV/AIDS than white individuals,” said Smith, which is why “initiatives like that in D.C. will be very important.”
Intermittent PrEP
While randomized trials and feasibility studies of test and treat have only recently started, several clinical trials of PrEP will soon be yielding results. Kenneth Mayer, an investigator involved in a PrEP trial in the US, said 2010 will be a “major year in our understanding of PrEP.” The results of four oral PrEP trials and one trial of topical PrEP (a microbicide gel formulation of tenofovir) are expected this year.
Two new PrEP studies (HPTN 066 and 067) are also slated to begin this year. These trials are evaluating the intermittent use of PrEP, rather than daily. “Intermittent PrEP dosing may be a more feasible approach for many populations,” said J. Gerardo Garcia-Lerma, a researcher at the US Centers for Disease Control and Prevention. The preliminary results of a pilot feasibility study of intermittent PrEP conducted by IAVI are also expected this year. This study, in addition to HPTN 066 and 067, which will collect extensive
pharmacokinetic data, will hopefully shed light on the optimal dosing regimens for intermittent PrEP.
At last year’s CROI, Garcia-Lerma presented results from an intermittent PrEP study in rhesus macaques that showed that when the drug Truvada (a combination of tenofovir and emtricitabine or FTC) was administered two hours before or after a low-dose rectal SHIV challenge, there was a four-fold reduction in risk of infection compared to untreated control animals (10). If Truvada was administered either one, three, or seven days prior to SHIV exposure with a post-exposure dose two hours after challenge, there was also a high level of protection, with a reduction in risk of infection ranging from 9- to 17-fold compared to untreated controls. “If the drug was administered close to the time of exposure, there was signification protection,” said Garcia-Lerma.
To better understand the relationship between this protection and the drug levels of Truvada, Garcia-Lerma and colleagues conducted single-dose pharmacokinetic studies in rhesus macaques. Drug concentrations in four animals were measured in plasma, peripheral blood mononuclear cells (PBMCs), and rectal secretions, and necropsy was used to collect lymphoid and rectal tissues from seven additional animals to determine tenofovir concentrations at these sites.
The results indicate that the concentration of the two drugs varies greatly between different sites. Garcia-Lerma said that both tenofovir and FTC persisted for a long time in PBMCs—the half-life of tenofovir diphosphate was about five days in these cells, similar to what is observed in humans, and the half-life for FTC was one day. In rectal secretions, however, the concentration of both drugs peaked at 24 hours. And, in plasma, drug levels of both tenofovir and FTC peaked after just two hours and then declined over one day. These findings suggest that the timing of intermittent PrEP may be important. “We have many things to learn about optimal [PrEP] dosing,” said Mayer.
Garcia-Lerma also presented data from a new study in rhesus macaques that shows that a single dose of PrEP is not nearly as effective at protecting against a low-dose, repeat SHIV challenge. “The post-exposure dose appears to be essential for protection,” concluded Garcia-Lerma.
The study compared two groups of six monkeys. One group received a single dose of Truvada three days before challenge, while the other received a single dose of the tenofovir pro-drug, GS7340. In Phase I and II clinical trials, GS7340 resulted in 100-fold higher tenofovir concentrations in PBMCs 24 hours after dosing as compared to orally administered tenofovir. All animals, including six controls, were exposed to a low-dose SHIV challenge once a week.
Five of the six macaques that received only a single dose of Truvada became infected after seven challenges, at which time the experiment was stopped. GS7340 didn’t fare any better—four of the six animals were infected after four challenges. “This is quite unexpected because of the high drug levels in PBMCs,” said Garcia-Lerma.
When researchers analyzed the GS7340 drug levels, they found no difference between the animals that became infected and those who didn’t. They also found that the concentration of the prodrug in PBMCs persisted for a long time—levels at 17-21 days after dosing were still higher than what is seen following oral administration of tenofovir. High drug levels also were detected in both lymph node and rectal tissues. Additionally, the acute viremia in GS7340-treated animals who became infected was blunted compared to controls, which Garcia-Lerma said was “another indication that the drug was having an effect.”
Yet despite all of this, the drug was unable to protect in this model. Garcia-Lerma said there were many factors that might explain this, including potential differences between the drug’s activity at effector and inductor immune sites at the mucosa. He suggested that evaluating the pharmacodynamics of different dosing schedules will be important for determining the most effective PrEP strategy.
Topical PrEP
In addition to oral dosing, researchers are also studying the use of gel formulation of ARVs that can be used as microbicides. Several efficacy trials of other non-ARV-based microbicides have provided disappointing results, with the latest in a string of failures reported in December 2009 (see Vaccine Briefs, IAVI Report, Nov.-Dec. 2009). “Microbicides without ARVs aren’t a dead issue, but they’re definitely on a resuscitator,” said Fauci. “The time has come to look at an ARV-based microbicide.”
There are two clinical trials underway to evaluate the use of either tenofovir or Truvada as a topical microbicide, with the first results expected later this year. Researchers are also experimenting with other ARVs that may be effective topical PrEP agents. One of these is maraviroc, the first licensed ARV that blocks HIV entry into cells by binding to the CCR5 receptor. Maraviroc was licensed for the treatment of HIV infection by Pfizer, who granted a license for development of the drug as a vaginal microbicide to the International Partnership for Microbicides (IPM).
Neither IPM nor Pfizer planned to evaluate the efficacy of maraviroc in nonhuman primates, according to John Moore, 
a professor of microbiology and immunology at Weill Cornell Medical College. So, he and his colleagues at the Tulane National Primate Research Center decided to take matters into their own hands. “Success in a monkey model doesn’t guarantee it will work in women but if you fail in the monkey, it’s not a good sign,” said Moore.
After crushing maraviroc pills and dissolving them in saline to prepare a make-shift microbicide, Moore and colleagues evaluated the efficacy of the vaginally applied compound to protect against a single, high-dose vaginal SHIV162P3 challenge in rhesus macaques, which were previously treated with progesterone to thin their vaginal epithelium.
They found that maraviroc was able to protect the monkeys against this stringent challenge in a dose-dependent manner. Half of the animals were protected at a maraviroc dose of 0.25 mg/ml. The fully protective dose was 4 ml at a concentration of 3 mg/ml. Moore reported that protection also waned over time—the longer the time between application and challenge, the less protective the drug was. The half-life of protection was approximately four hours.
Since a CCR5 inhibitor would obviously not protect against virus that utilizes the CXCR4 receptor to gain entry into cells, Moore and colleagues also conducted a control experiment in which they challenged maraviroc-treated macaques with a high-dose X4-tropic SHIV. As expected, all the animals were infected, but the viral loads in X4-infected animals were the same as in the untreated controls, suggesting to Moore that there was “no indication that maraviroc exacerbates X4 infection.”
* Richard Jefferys is Basic Science Project Director at the Treatment Action Group, a New York-based organization advocating for HIV research.
1. N. Engl. J. Med. 361, 2209, 2009
2. Nat. Med. 15, 951, 2009
3. Proc. Natl. Acad. Sci. 102, 18514, 2005
4. N. Engl. J. Med. 345, 715, 2001
5. Proc. Natl. Acad. Sci. 106, 20234, 2009
6. Science 308, 1906, 2005
7. Molecules 14, 4892, 2009
8. PLoS Pathog. 5, e1000272, 2009
9. Lancet 373, 48, 2009
10. Sci. Transl. Med. 2, 14ra4, 2010